Low-power VLSI decoder architectures for LDPC codes
Proceedings of the 2002 international symposium on Low power electronics and design
Errata for "good error-correcting codes based on very sparse matrices"
IEEE Transactions on Information Theory
| Hi-index | 0.01 |
This paper proposes the power-efficient LDPC decoder architecture which features (1) a FIFO buffering based rapid convergence schedule which enables the decoder to accelerate the decoding throughput without increasing the required number of memory bits, (2) an intermediate message compression technique based on a clock gated shift register which reduces the read and write power dissipation for the intermediate messages. Simulation results show that the proposed decoder achieves 1.66 times faster decoding throughput, and improves the power efficiency (which is defined by the power dissipation per Mbps) up to 52% compared to the decoder based on the conventional overlapped schedule.